A Powerful Tool For Model Evaluation

Get to Know the Matthews Correlation Coefficient (MCC)

A Powerful Tool for Model Evaluation

What is MCC?

The Matthews correlation coefficient (MCC) is a statistical measure that assesses the performance of a classification model. It was invented by Brian Matthews in 1975 and is used to evaluate the ability of a classifier to distinguish between different classes.

Why Use MCC?

MCC is a robust measure that takes into account both true positives and true negatives, making it a reliable indicator of a model's overall performance. It is particularly useful for imbalanced datasets, where there are significant differences in the number of observations in each class.

How to Calculate MCC?

The formula for calculating MCC is:

``` MCC = (TP * TN - FP * FN) / sqrt((TP + FP) * (TP + FN) * (TN + FP) * (TN + FN)) ``` where: * TP: True positives * TN: True negatives * FP: False positives * FN: False negatives

Example

In a dataset with two classes (positive and negative), a model correctly classifies 100 positive observations and 90 negative observations. It misclassifies 10 positive observations as negative and 5 negative observations as positive. The MCC for this model is:

``` MCC = (100 * 90 - 10 * 5) / sqrt((100 + 10) * (100 + 5) * (90 + 10) * (90 + 5)) = 0.95 ``` This indicates that the model has a high level of discrimination between the two classes.